Wednesday, March 26, 2014

Look for segments in the youth vote

It is easy to get sloppy when we talk about specific groups of voters. Mention Asian American voter behavior and you have to differentiate between Chinese, Hmong, Indian, Filipino, Vietnamese, and other types of voters. Talk about the Hispanic vote and you have certain Hispanic voters who also identify as white, Hispanic voters whose families have lived in the United States for a varied number of generations, and the same national cleavages mentioned previously (Cuban, Mexican, Dominican, Puerto Rican, etc.).

So it should come as no surprise that we face these same distinctions within the all-encompassing 'youth vote.' This article proposes that the segmentation of the electorate is more important than aggregating swathes of voters, particularly in regards to the youth vote.

Slate recently caught the New York Times with its pants down when the esteemed newspaper discussed 'millennials.' The problem was that the Times managed to write several stories about millennials without talking to millennials.
According to the Times itself, Neil Howe and William Strauss—the men who literally wrote the book on millennials and are credited with coining the term—establish the start of the millennial generation with people born in the year 1982. That means that today, even the eldest millennials are no more than 32 years old. And yet, the Times trend story on millennial mourning quotes Soffer, the 37-year-old founder of online grief resource Modern Loss; her co-founder Gabrielle Birkner (at 34, not a millennial); 35-year-old Modern Loss blogger Melissa Lafsky Wall (not a millennial); and Jason Feifer, the 33-year-old creator of the Tumblr “Selfies at Funerals” (so close, and yet, not a millennial).
The kicker comes at the end of the piece.
When millennial trend stories stick to their self-imposed age limits, they produce annoying stereotypes about a group of 80 million Americans. But when reporters can’t even find enough of us to fill out their pieces, they’re just lying. Trend stories are anecdotes in search of a generalization, and choosing the organizing principle of “millennial” allows the Times to pretend that it’s really reporting on something new.  
Think about this for a minute. When we talk about the youth vote, folks often get lazy with their terminology. "Millennials" include people age 32 on down. Is the cutoff age 15? 18? We are potentially talking something like a 20 year span of individuals. Exit polling often counts the youth vote as aged 18-29. Since 2006, exit polls have begun to dig deeper, showing us numbers for voters aged 18-24 as well. Which segment of the youth vote should we be targeting?

Nobody should be surprised that voters aged 18-24 vote somewhat differently from those aged 18-29. Even within the aggregate monstrosity we call the youth vote, segments of young Americans vote differently than each other.


For simplicity, there are two schools of thought among consultants targeting the youth vote. A good consultant will draw on both methodological traditions to maximize their effectiveness, but let's assume you are limited to one (the Obama campaign was not limited to one, and reading The Victory Lab by Sasha Issenberg is a great case study to see those efforts in action).

However, the two competing traditions for the youth vote can be summed up as field heavy and digital heavy. I wrote about the field heavy version at Red Racing Horses (RRH):
When we talk about voter contact... I will generally pick direct voter contact [field], and in particular I would pick door-to-door. Study after study shows that it is an effective strategy for mobilizing and persuading voters (see Green and Gerber's Get Out The Vote) by closely approximating real personal interaction, and the only real limits to using it are the time and labor costs involved (financing a statewide canvassing operation is expensive!)...
The flip-side of this coin is indirect voter contact, which among young voters often means digital advertising or voter contact through social media platforms like Facebook. My RRH colleague, Kevin Palmer, describes this approach well at Red Alert Politics.
Ken Cuccinelli’s unsuccessful campaign for governor of Virginia last fall was generally regarded as a subpar effort, but the Republican shockingly managed to win the 18-24 demographic by 6 points despite losing the overall election by 2. Cuccinelli’s breakthrough with Virginia’s youngest voters may have been assisted by the College Republican National Committee, which ran ads in his support on Hulu, YouTube and Pandora, all of which are disproportionately used by young people. By reaching young people on platforms they use, and speaking a language unique to them (the ads were in the style of MTV’s popular Catfish show), the CRNC was able to move elusive voters for a much lower sum of money than it would have cost to run the ads on TV or radio.
This approach stresses finding young voters online and prioritizes financing digital on the campaign to reach young voters (my post on the youth vote at RRH also looked at this Virginia gubernatorial election, and I happen to fall more on the field side of this dichotomy).


I want to take a step back from that debate and point out an obvious reason why this dichotomy may be short-sighted. Certain segments of the youth vote may respond better to field than to digital efforts, and vice-versa. For example, a field campaign targeting the segment of college students clumped in a dorm is much easier to conduct than attempting to locate the physical addresses of non-students aged 18-24, who may be difficult to find if they have left home. Instead, those voters may require a digital touch, since similar to their college counterparts they likely use social media or sites like Youtube or Hulu.

Similarly, age segments of the youth vote probably respond differently to different campaign treatments. Who expects a 32 year-old born in 1982 to vote the same way as an 18 year-old born in 1995? Both are millennials, but their voting patterns are educated by wildly different experiences. Back in the 1980s, the cool thing for young people to do was vote Republican. Those voters grew up with Ronald Reagan and voted accordingly.

The Monkey Cage sees a similar phenomenon among two different segments of the youth vote.
Those who came of age during Clinton’s second term or Bush’s second term were more likely to vote for Obama, to identify with or lean toward the Democratic Party, to identify as liberal, and to agree that “it is the responsibility of the federal government to see to it that everyone has health care coverage” — relative to those who came of age during Bush’s first term or Obama’s first term.  In fact, those millennials who came of age during Obama’s first term were the least likely of these four groups to vote for him or identify as Democratic.

If we zero in even further on the youngest of the millennials in these polls — those who turned 18 during Obama’s first term — the potential challenges for Democrats become even clearer. Among self-reported voters who were 18 years old in 2012, Mitt Romney, not Obama, won the majority: 57 percent.  Romney also won 59 percent among 19-year-olds, and 54 percent among 20-year-olds.  These youngest voters of 2012 had entered the electorate in 2010-2012, when Obama’s popularity was much lower than the high point of his inauguration.  Only among “the oldest of the youngest” — 21-year-olds, whose political memories would have been forged during Obama’s first year in office and perhaps during his first presidential campaign — did Obama win a clear majority (75 percent).
If we imagine the youth vote consists of numerous segments, those aged 18-20 are the most anti-Obama, while those aged 21-24 are more in favor of the President. Or digging into those Virginia exit polls, voters aged 18-24 voted for Cuccinelli while those aged 18-29 favored McAuliffe. We can constitute and reconstitute the youth vote into wider or narrower segments and see different results. As one commentator notes,
18 to 20 year olds are a diverse lot. White, Black and Hispanic young voters are likely to vote very differently, whereas those who leave home for four year university are also likely to have an entirely different behavioural output than those who went to work full-time after secondary school.

In sum, politics is messy from an analytical standpoint. Every time we think we have wrapped our heads around a meaningful group of voters, a lurking variable can throw all of our careful analysis out the window. We can slice and dice the electorate by age, by education, or by race, and each time we will discover a different output for voter behavior.

The problem is particularly acute with young voters, who are difficult to locate and contact. Without thorough studies quantifying and diving into the data, our real understanding of young Americans' voting patterns remains limited. So avoid broad explanations of what millennials want or how they behave. You will find more meaning exploring segments of the youth vote than in the aggregate mass of millennials.

No comments:

Post a Comment